PROJECT SUMMARY/ABSTRACT I am applying for a Career Development Award (K23) in ocular oncology. A K23 will allow me to establish a clinical research program to enhance prognostication and improve holistic care for patients with choroidal melanoma. The proposed research and training will provide the foundation for an R01/UG1 proposal to develop an international ocular melanoma registry, fostering multi-center comparative studies with large datasets. My long-term career goals are to: 1) lead a multidisciplinary ocular oncology team at University of California, San Francisco (UCSF); 2) create a prospective database of clinical outcomes, molecular pathology data and patient- reported outcomes (PROs) to predict ocular outcomes, survival and quality of life (QOL); 3) advance randomized trials and multicenter collaboration in ocular oncology, to advance diagnostics and therapeutics. My training objectives are to: 1) master advanced biostatistics methodology, database management and study design through didactic courses and apprenticeships, working with renowned mentors in ocular oncology, molecular pathology and statistics; 2) understand molecular diagnostics in uveal melanoma and how to evaluate new prognostic and diagnostic tests in clinical practice, through mentored laboratory experiences with a pre-eminent molecular pathologist; 3) engage in multidisciplinary research, in collaboration with international colleagues, forming connections for future multi-center studies; and 4) become proficient in scientific writing. The specific aims of this project are to: 1) develop and validate prognostic models for predicting visual outcome, local tumor control and ocular conservation after proton beam radiotherapy for choroidal melanoma. We will test the hypothesis that radiation dosimetry predicts such outcomes better than baseline clinical features. 2): validate a targeted next-generation genetic sequencing (NGS) assay as a prognostic test for survival in choroidal melanoma. We hypothesize that NGS of choroidal melanoma reliably determines tumor lethality. 3): develop prognostic models for predicting PROs and QOL after proton beam radiotherapy for choroidal melanoma. We hypothesize that social factors are more predictive than ophthalmic features in determining PROs and QOL We will perform secondary analysis of an existing dataset of patients treated at the Liverpool Ocular Oncology Centre, mostly by the primary mentor (BD), to create and validate the prognostic models, which we will then evaluate in a cohort of patients at UCSF.